Skip to content

Commit 5017eda

Browse files
authored
[Enhancement] modified template to automatically generate resources.claim (#195)
* modified template to automatically generate resources.claim Signed-off-by: ryojsb <ryoh80213@gmail.com> * modified chart version Signed-off-by: ryojsb <ryoh80213@gmail.com> * modified chart version on example manifest Signed-off-by: ryojsb <ryoh80213@gmail.com> * removed unnecessary parameter from values.yaml Signed-off-by: ryojsb <ryoh80213@gmail.com> --------- Signed-off-by: ryojsb <ryoh80213@gmail.com>
1 parent c97db00 commit 5017eda

21 files changed

Lines changed: 23 additions & 88 deletions

charts/llm-d-modelservice/templates/_helpers-dra.tpl

Lines changed: 8 additions & 22 deletions
Original file line numberDiff line numberDiff line change
@@ -112,8 +112,8 @@ false
112112
{{- $count -}}
113113
{{- end }}
114114

115-
{{/* Generate pod-level resourceClaims (merges accelerator + user-defined claims) */}}
116-
{{- define "llm-d-modelservice.podResourceClaims" -}}
115+
{{/* Generate resourceClaims Variable (merges accelerator + user-defined claims) */}}
116+
{{- define "llm-d-modelservice.resourceClaimsBase" -}}
117117
{{- $claims := list -}}
118118
{{- $draEnabled := eq (include "llm-d-modelservice.draEnabled" .) "true" -}}
119119
{{- if $draEnabled -}}
@@ -125,29 +125,15 @@ false
125125
{{- $claims = concat $claims .pdSpec.resourceClaims -}}
126126
{{- end -}}
127127
{{- if $claims -}}
128-
resourceClaims:
129-
{{- toYaml $claims | nindent 2 }}
128+
{{- toYaml $claims }}
130129
{{- end -}}
131130
{{- end }}
132131

133-
{{/* Generate container-level resource claims (merges accelerator + user-defined claims) */}}
134-
{{- define "llm-d-modelservice.containerResourceClaims" -}}
135-
{{- $claims := list -}}
136-
{{- $draEnabled := eq (include "llm-d-modelservice.draEnabled" .) "true" -}}
137-
{{- if $draEnabled -}}
138-
{{- $claimName := include "llm-d-modelservice.acceleratorClaimName" . -}}
139-
{{- $claims = append $claims (dict "name" $claimName) -}}
140-
{{- end -}}
141-
{{- if and .resources .resources.claims -}}
142-
{{- if kindIs "slice" .resources.claims -}}
143-
{{- $claims = concat $claims .resources.claims -}}
144-
{{- else -}}
145-
{{- fail "resources.claims must be a list of objects with 'name' field, e.g., [{\"name\": \"claim-name\"}]" -}}
146-
{{- end -}}
147-
{{- end -}}
148-
{{- if $claims -}}
149-
claims:
150-
{{- toYaml $claims | nindent 2 }}
132+
{{- define "llm-d-modelservice.podResourceClaims" -}}
133+
{{- $claimList := include "llm-d-modelservice.resourceClaimsBase" . -}}
134+
{{- if $claimList -}}
135+
resourceClaims:
136+
{{- $claimList | nindent 2 }}
151137
{{- end -}}
152138
{{- end }}
153139

charts/llm-d-modelservice/templates/_helpers.tpl

Lines changed: 10 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -281,15 +281,14 @@ nvidia.com/gpu
281281
{{- $requests = deepCopy .resources.requests }}
282282
{{- end }}
283283
{{- $draEnabled := eq (include "llm-d-modelservice.draEnabled" .) "true" -}}
284+
resources:
284285
{{- if $draEnabled -}}
285286
{{- /* DRA mode: pass through user-defined limits/requests as-is, add claims */}}
286287
{{- /* Users should not include accelerator resources in limits when DRA is enabled */}}
287-
resources:
288288
limits:
289289
{{- toYaml $limits | nindent 4 }}
290290
requests:
291291
{{- toYaml $requests | nindent 4 }}
292-
{{- include "llm-d-modelservice.containerResourceClaims" . | nindent 2 }}
293292
{{- else -}}
294293
{{- /* Device Plugin mode: existing logic */}}
295294
{{- $numGpus := int (include "llm-d-modelservice.numGpuPerWorker" .parallelism) -}}
@@ -307,17 +306,20 @@ resources:
307306
{{- $requests = mergeOverwrite $requests (dict $acceleratorResource (toString $numGpus)) }}
308307
{{- end }}
309308
{{- end }}
310-
resources:
311309
limits:
312310
{{- toYaml $limits | nindent 4 }}
313311
requests:
314312
{{- toYaml $requests | nindent 4 }}
315-
{{- /* Include user-defined claims even in Device Plugin mode */}}
316-
{{- if and .resources .resources.claims }}
313+
{{- end -}}
314+
{{- $claimList := include "llm-d-modelservice.resourceClaimsBase" . | fromYamlArray -}}
315+
{{- if $claimList }}
317316
claims:
318-
{{- toYaml .resources.claims | nindent 4 }}
317+
{{- $containerClaims := list -}}
318+
{{- range $claimList -}}
319+
{{- $containerClaims = append $containerClaims (dict "name" .name) -}}
319320
{{- end }}
320-
{{- end -}}
321+
{{- toYaml $containerClaims | nindent 4 }}
322+
{{- end }}
321323
{{- end }}
322324

323325
{{/* prefill name */}}
@@ -494,7 +496,7 @@ context is a dict with helm root context plus:
494496
startupProbe:
495497
{{- toYaml . | nindent 4 }}
496498
{{- end }}
497-
{{- (include "llm-d-modelservice.resources" (dict "resources" .container.resources "parallelism" .parallelism "container" .container "Values" .Values "role" .role)) | nindent 2 }}
499+
{{- (include "llm-d-modelservice.resources" (dict "resources" .container.resources "parallelism" .parallelism "container" .container "Values" .Values "role" .role "pdSpec" .pdSpec)) | nindent 2 }}
498500
{{- include "llm-d-modelservice.mountModelVolumeVolumeMounts" (dict "container" .container "Values" .Values) | nindent 2 }}
499501
{{- /* DEPRECATED; use extraConfig.workingDir instead */ -}}
500502
{{- with .container.workingDir }}

charts/llm-d-modelservice/templates/decode-deployment.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -56,7 +56,7 @@ spec:
5656
{{- with .Values.decode.containers }}
5757
containers:
5858
{{- range . }}
59-
{{- (include "llm-d-modelservice.container" (dict "role" "decode" "container" . "parallelism" $.Values.decode.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart)) | nindent 8 }}
59+
{{- (include "llm-d-modelservice.container" (dict "role" "decode" "container" . "parallelism" $.Values.decode.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart "pdSpec" $.Values.decode)) | nindent 8 }}
6060
{{- end }}
6161
{{- end }}
6262
{{- if .Values.decode.tolerations }}

charts/llm-d-modelservice/templates/decode-lws.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -74,7 +74,7 @@ spec:
7474
{{- with .Values.decode.containers }}
7575
containers:
7676
{{- range . }}
77-
{{- (include "llm-d-modelservice.container" (dict "role" "decode" "container" . "parallelism" $.Values.decode.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart)) | nindent 8 }}
77+
{{- (include "llm-d-modelservice.container" (dict "role" "decode" "container" . "parallelism" $.Values.decode.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart "pdSpec" $.Values.decode)) | nindent 8 }}
7878
{{- end }}
7979
{{- end }}
8080
{{- if .Values.decode.nodeSelector }}

charts/llm-d-modelservice/templates/decode-requester-replicaset.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -34,7 +34,7 @@ spec:
3434
{{- with .Values.decode.containers }}
3535
containers:
3636
{{- range . }}
37-
{{- (include "llm-d-modelservice.container" (dict "role" "decode" "container" . "parallelism" $.Values.decode.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart)) | nindent 14 }}
37+
{{- (include "llm-d-modelservice.container" (dict "role" "decode" "container" . "parallelism" $.Values.decode.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart "pdSpec" $.Values.decode)) | nindent 14 }}
3838
{{- end }}
3939
{{- end }}
4040
spec:

charts/llm-d-modelservice/templates/prefill-deployment.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -51,7 +51,7 @@ spec:
5151
{{- with .Values.prefill.containers }}
5252
containers:
5353
{{- range . }}
54-
{{- (include "llm-d-modelservice.container" (dict "role" "prefill" "container" . "parallelism" $.Values.prefill.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart)) | nindent 8 }}
54+
{{- (include "llm-d-modelservice.container" (dict "role" "prefill" "container" . "parallelism" $.Values.prefill.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart "pdSpec" $.Values.prefill)) | nindent 8 }}
5555
{{- end }}
5656
{{- end }}
5757
{{- if .Values.prefill.tolerations }}

charts/llm-d-modelservice/templates/prefill-lws.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -69,7 +69,7 @@ spec:
6969
{{- with .Values.prefill.containers }}
7070
containers:
7171
{{- range . }}
72-
{{- (include "llm-d-modelservice.container" (dict "role" "prefill" "container" . "parallelism" $.Values.prefill.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart)) | nindent 8 }}
72+
{{- (include "llm-d-modelservice.container" (dict "role" "prefill" "container" . "parallelism" $.Values.prefill.parallelism "Values" $.Values "Release" $.Release "Chart" $.Chart "pdSpec" $.Values.prefill)) | nindent 8 }}
7373
{{- end }}
7474
{{- end }}
7575
{{- if .Values.prefill.nodeSelector }}

charts/llm-d-modelservice/values.yaml

Lines changed: 0 additions & 28 deletions
Original file line numberDiff line numberDiff line change
@@ -347,20 +347,6 @@ decode:
347347
# additionalProperties: true
348348
# @schema
349349
requests: {}
350-
# @schema
351-
# type: array
352-
# items:
353-
# type: object
354-
# required: [name]
355-
# properties:
356-
# name:
357-
# type: string
358-
# @schema
359-
# Resource claims for this container (DRA)
360-
# Will be merged with accelerator claim when accelerator.dra is enabled
361-
claims: []
362-
# Example:
363-
# - name: llm-d-imex-channel-0
364350
# when set, a volumeMount (and volume) is created for model storage
365351
mountModelVolume: true
366352
# @schema
@@ -579,20 +565,6 @@ prefill:
579565
# additionalProperties: true
580566
# @schema
581567
requests: {}
582-
# @schema
583-
# type: array
584-
# items:
585-
# type: object
586-
# required: [name]
587-
# properties:
588-
# name:
589-
# type: string
590-
# @schema
591-
# Resource claims for this container (DRA)
592-
# Will be merged with accelerator claim when accelerator.dra is enabled
593-
claims: []
594-
# Example:
595-
# - name: llm-d-imex-channel-0
596568
# @schema
597569
# type: array
598570
# items:

examples/output-cpu.yaml

Lines changed: 0 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -89,7 +89,6 @@ spec:
8989
ports:
9090
- containerPort: 8200
9191
protocol: TCP
92-
9392
resources:
9493
limits:
9594
{}
@@ -161,7 +160,6 @@ spec:
161160
ports:
162161
- containerPort: 8000
163162
protocol: TCP
164-
165163
resources:
166164
limits:
167165
{}

examples/output-dra.yaml

Lines changed: 0 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -94,7 +94,6 @@ spec:
9494
name: llm-d-hf-token
9595
key: HF_TOKEN
9696

97-
9897
resources:
9998
limits:
10099
cpu: "4"

0 commit comments

Comments
 (0)